Purpose
The methodology is designed to make the limits of the catalogue visible. It does not award products a universal quality mark. It records what is known, where that information came from and when it was last checked.
Evidence labels
- Verified
- A specific claim has been checked against an appropriate source and should carry a verification date where it can change.
- Supplier stated
- The supplier makes the claim, but GP Trusted Tools has not independently verified it.
- Needs evidence review
- The catalogue entry is retained, but current product details have not yet been checked.
How a claim is assessed
- Identify the precise claim rather than relying on broad marketing language.
- Choose a source appropriate to that type of claim.
- Record any important qualification, territory or product-version limitation.
- Date information that may change.
- Use a cautious label when the available source does not justify a stronger statement.
What counts as evidence
The appropriate source depends on the claim. Official product documentation may support a feature description. An authoritative framework owner, regulator or standards body is needed for claims about procurement routes, regulatory status, accreditation or formal compliance. Independent research may provide useful evidence about outcomes, but its population, setting and limitations must be considered.
Customer logos, testimonials, marketplace listings and observed use may show that a product is used somewhere; they do not by themselves establish NHS endorsement, clinical validation, interoperability or suitability for another practice.
Product information and reviews are different
Catalogue facts describe the product and its documented position. Reviews describe a contributor’s experience in a particular context. A review does not verify a supplier’s general claim, and a verified product fact does not prove that every implementation will produce the same experience.
Time-sensitive information
Pricing, integrations, availability, ownership, procurement routes and product functionality can change. These should be dated and rechecked. Where no reliable current source is available, the site should say so rather than infer an answer.
Comparisons and missing information
Comparison tables should use common fields and preserve unknown values. A blank or unknown entry must not be treated as evidence that a feature is absent. Differences in evidence quality should remain visible rather than being converted into an artificial score.
Corrections
Material corrections should preserve the distinction between sourced facts, supplier statements, editorial interpretation and moderated reviewer experience. The repository history provides an auditable record of changes. To suggest a correction, use the details on the contact page and include a source where possible.